The proposed research will test the following hypotheses: Hypothesis 1: The incidence of IDDM in Allegheny County will continue to rise. Hypothesis 2: There is major geographic variation in the United States and North America by race and ethnic group. Hypothesis 3: There are epidemics of IDDM in our hemisphere and these are associated with the occurrence of chicken pox. Hypothesis 4: High risk HLA DQ heterodimers are independent determinants of diabetes in first degree relatives of probands having IDDM, particularly in the presence of islet cell or GAD antibodies. To evaluate hypotheses 1-3 we will assess the incidence of childhood diabetes in standardized population based registries from Allegheny County, PA plus 17 other directly comparable registries from North America. To test hypothesis 4 we will continue our 15 year follow-up of over 4000 first degree relatives of IDDM cases to examine what precedes overt insulin-dependent diabetes mellitus.